The Polymerization of Styrene in the Presence of Nitrobenzene, 2,4

Charles C. Price, Dorothy Ann Durham. J. Am. Chem. Soc. , 1943, 65 (5), ... Walter H. Stockmayer , Leighton H. Peebles Jr. Journal of the American Che...
0 downloads 0 Views 313KB Size
May, 1943

THEPOLYMERIZATION OF STYRENE

that melted at 160-161.5° after crystallization from benzene-ligroin. Anal. Calcd. for CUHISO~:C, 81.8; H, 4.6. Found: C,81.7; H,4.6. 2-Acetoxy-J,5,6-~phenplquinone, prepared from the hydroxyquinone with acetic anhydride and sulfuric acid, formed orange needles that melted at 185.5-187.5’ after crystallization from acetic acid. Anal. Calcd. for cdb~o4:c, 79.2; H,4 6. Found: C, 79.5; H, 4.5. The acetate was readily saponified by alcoholicpotassium hydroxide.

[CONTRIBUTION FROM TEE

757

summary 2,4,4-Triphenylcyclopentenedione-1,3 is readily alkylated by ethyl bromoacetate, and when the resulting diketoester is treated with sodium ethoxide it undergoes rearrangement to ethyl 3,4,6triphenylgentisate, A number of transformations of this rearrangement product are described, irlcluding its into tfiphenylhyhoquinone and hyhoxfifiphenylquinone. MINNEAPOLIS,

MINNESOTARECEIVED JANUARY 30, 1943

NOYES CHEMICAL LABORATORY, UNIVERSITY OF

ILLINOIS]

The Polymerization of Styrene in the Presence of Nitrobenzene, 2,4-Dinitrochlorobenzene and Nitromethane B Y CHARLES

c. PRICE AND DOROTHY ANN DURHAM

The few brief observations‘ that certain vinyltype polymerizations are inhibited or retarded by various aromatic nitro compounds have attracted very little attention. There has been no direct experimental evidence as to the specific nature of the effect and no explanation has been suggested to account for this very interesting phenomenon. Since the polymerizations which are subject to this effect of aromatic nitro compounds are those which proceed through an intermediate active free radical,‘c’’ there exists the possibility that the active radical may become inactivated by coupling with the aromatic nitro compound. The preparation of polystyrene in the presence of nitrobenzene and 2,4-dinitrochlorobenzene has confirmed this surmise since the samples contained the elements of the nitro compound in a ratio as high as the theoretical ratio of one nitroaryl residue per polymer molecule. Several samples of polystyrene prepared in the presence of a small amount of 2,4dinitrochlorobenzene contained chlorine (2.3 to 3.273 and nitrogen (2.3 to 2.7%) corresponding to one dinitrochlorophenyl group per polymer molecule. A sample of polystyrene prepared in the presence of nitrobenzene contained nitrogen (0..53~o)corre(1) (a) Ostromislensky and Shepard, U. S. Patent 1,550,323, August 18, 1925; (b) Foord, J . Chcm. Soc., 48 (1940), (c) Price, Kell and Krebs, THISJOURNAL, 64, 1103 (1942). (2) Staudinger, Trans. Faraday Soc., 89, 323 (1930); Flory, THIS JOURNAL, 69, 241, (1937); Schulz and Wittig, Nafurwissendmflcn, 27, 387, 456 (1939); Schulz, ibid., 659 (1939); Norrisa and Brookman, Proc. ROY. SOC.(London), A171, 147 (1939); Norrish, Trans. Faraday SOC.,86, 1087 (1939); Kamenskaya and Medvedev, Acta Phyricochnn., U.S.S . R., 18, 565 (1940); Melville, Proc. Roy. SOC.(London), A168, 511 (1937).

sponding to one nitrophenyl group for every other polymer molecule. Several other samples of polystyrene prepared in the presence of smaller amounts of these two nitro compounds also contained the elements of the nitro compound but in an appreciably smaller proportion than the maximum of one per polymer molecule. These results may be accounted for on the basis of competition between the reaction of the growing chain with retarder and the normal cessation process. The chance for normal cessation to occur would, of course, be favored by decreased concentration of the retarder. ‘The most logical and satisfactory manner to account for the destruction of the activity of the growing free radical polymer chain by reaction with a molecule is by a process which will transform the “active” radical into an “inactive” one3: the odd electron characteristic of the free radical may only be “neutralized” by pairing with the odd electron of another free radical. On this basis, the observed coupling of the growing chain with aromatic nitro compounds must involve an intermediate addition compound in which the odd electron has become sufficiently stabilized so that it will no longer add t o monomer molecules. Since coupling of free radicals with nitrobenzene takes place principally in the paraposition, the over-all scheme for the benzoyl peroxide-catalyzed polymerization of styrene in the piesence of nitrobenzene may be indicated as follows (3) The terms “active” and “inactive” aere refer solely to whether or not the radicals are reactive enough to add to a monomer molecule and propagate the polymerization.

758

CHARLES C. PRICE AND DOROTHY ANN DURHAM

Vol. 65

That the effect of aromatic nitro comportnds in retarding free radical polymerization is indeed due to an activation of the aromatic nucleus by the nitro groups and not to a direct reaction of the CsHs C"H& nitro group itself was substantiated by carrying out the polymerization of styrene in nitromethane. The polymerization proceeded a t the normal rate to give a good yield of polymer which contained no detectable amount of nitrogen. The inert character of the nitro group itself toward free radicals was further indicated by the A plausible explanation for the effect of the isolation of diphenyl from the decomposition of nitro group in promoting such a reaction of the benzoyl peroxide in nitromethane. Evidently, benzene ring with the growing free radical chain,4 this solvent is sufficiently inert to allow the highlyas well as in stabilizing the intermediate addireactive phenyl free radicals, formed as intermetion product (A), is the increased number of possidiates in the reaction, a long enough life for an bilities for resonance stabilization in the interappreciable number to combine, yielding biphenyl. mediate A. Experimental' H Polystyrene in the Presence of 2,4-Dinitrochlorobenzene.-A solution of 5 g. of 2,4-dinitrochlorobenzeneand 5 g. of benzoyl peroxide in 25 cc. of styrene in a beaker was gradually warmed on a steam-bath. After ten to fifteen minutes the temperature of the reaction mixture reached 95" and then slowly fell to 86". After three hours, the This view is substantiated by the fact that an dark orange, slightly-viscous mixture was cooled, diluted increase in the number of nitro groups increases with ether and poured into alcohol. No precipitate the retarding influence.la The marked effect of formed, indicating that only very low molecular weight the possibility of resonance with the nitro group polystyrene was present. Addition of a small amount of in stabilizing a free radical is also indicated by water precipitated a very viscous, deep orange oil. It was redissolved in 10 cc. of ether and precipitated by pouriiig the work of Ziegler and BoyeS and of Allen and iuto 50 cc. of 90% alcohol. The creamy-white powdery Sugden.6 The former, working with 4,4',4"-tri- solid (I, 0.7 9.) was reprecipitated once more yielding 0.45 nitrotriphenylmethyl, and the latter, with a,a- g. for analysis. Anal. Calcd. for CsHbCO~(CeHe) d2~H~(NOS)~C : l-O~ diphenyl-0-trinitrophenylhydrazyl,have found these substances to be remarkably stable free C, 79.97; H, 6.29; N, 2.01; C1. 2.54; mol. wt., 1390. Found: C, 7924; H, 6.46; N, 2.26; C1, 2.33; mol. wt., radicals, existing as such in the solid state.6 1310 (visc.) H Addition of more water to the mother liquor from the precipitation of I gave another viscous oil. A second reprecipitation gave 1.4 g. of pale orange powder (11). On filtering a solution of 1 g. of this material in 10 cc. of ether into 50 cc. of filtered 90% alcohol, 0.4 g. of ivory-yellow material was obtained. Anal. Calcd. for CaHsCO*(C3Is)86H*(NOn)lCf.O: C. 78.92; H,6.11; N,2.39; C1,3.03; mol. wt., 1160. Found C, 79.02; H, 6.09; N, 2.444; C1, 2.84; mol. wt., 1180 (visc.).e Addition of water to the mother liquor from I1 gave a third sample (1.3 8.) of polymer. A solution of 1 g. of this material in 10 cc. of ether was filtered into 50 cc. of icecold 90% alcohol, yielding 0.25 g. of ivory-yellow powder (4) Since this manuscript was submitted this view haa been for analysis. The large loss on reprecipitation of these further substantiated by the observation of Fieser, Clapp and Daudt samples is an indication of the high percentage of very low (TXISJOURNAL, 64, 2052 (1942)) that nitro groups promote the -e of methylation of the benzene ring by means of the methyl radicals molecular weight polystyrene present. from acetyl peroxide, lead tetraacetate and electrolyzed scdium acetate. ( 6 ) Ziegler and Boye, A n n . , 458, 248 (1927) (6) Allen and Sugden, 3 Chcm Soc., 440, (1936)

(7) Analyses by Margaret McCarthy, Dorothy Schndder and Theta Spoor (8) Remp and Peters, Ind. Eng. C h c m . . 94, 1097 (1942).

Anal. Calcd. for CsHbCO2(CdH~)7CdH,(NOl)tC1.O:C. 77.62; H, 5.95; N, 2.02; C1,3.32; mol. wt., 1070. Found: C, 77.82; H, 0.07; N, 2.66; C1, 3.20; mol. wt., 1080 (visc.).' A second polymerization, identical with the one above except that only 1.0 g. of 2,4-dinitrochlorobenzene was added, gave 12 g. of pale orange powdery polymer on the first precipitation. A solution of 1 g. of this material in 10 cc. of ether was filtered into 50 cc. of 95% alcohol, yielding 0.8 g. of ivory white powder for analysis. A qualitative test by sodium fusion0 gave positive results for nitrogen and for halogen. Anal. Calcd. for CsHsCO&Ha)l, CsHs(CaH&,O CaHsCOz(C&Ia)laC+,H2(NO2)?Cl:C, 88.28; H, 7.21; N, 0.61; C1, 0.77; mol. wt. (average), 1530. Found: C, 88.10; H, 7.30; N, 0.47; C1,0.89; mol. wt., 1370 (vise.).' Polystyrene in the Presence of Nitrobenzene.-A solution of 3 cc. of nitrobenzene and 10 g. of benzoyl peroxide in 25 cc. of freshly-distilled styrene, was warmed gently a t 80-85" for twelve hours. Dioxane was added to the cooled reaction mixture and the polymer was precipitated by pouring into cold alcohol. After several reprecipitations, 6 g. of white powder was obtained. On sodium fusion, this material gave a positive qualitative test for nitrogen.' Anal. Calcd. for CbHsCOdCaHs)lz C"H,(CdH~)19CsHJVOz: C, 89.84; H, 7.36; N, 0.50; mol. wt., 1410. Found: C, 90.01; H, 7.33; N, 0.53; mol. wt., 1300 (visc.).' A second polymerization, using 5 cc. of nitrobenzene, 5 g. of benzoyl peroxide and 25 cc. of styrene was heated on the steam-bath in an evaporating dish. The temperature of the mixture rose to 100". After an hour at 90" the viscous orange oil was dissolved in ether and precipitated with alcohol yielding 5.5 g. of very pale orange powder. The clear yellow solution of 1 g. of th'ls material in 10 cc. of ether was filtered into 50 cc. of ice-cold 95% alcohol yielding a 0.75-9. sample for analysis. Anal. Calcd. for ~C~H~COZ(C&I&OOZGHbCOs(C&IS)&sH1NOp: C, 87.28; H, 7.12; N, 0.235; mol. wt., 1190. Found: C, 87.40; H, 7.07; N, 0.22, 0.28; mol. wt., 1130 (visc.).8 Polystyrene.-A blank experiment carried out under the same conditions as those above but omitting the retarder indicated its marked effect in moderating the vigor of the reaction and in decreasing the yield and molecular weight of the polymer formed. Actually, both the yield and molecular weight of polymer from this blank undoubtedly would have been still larger if it had been possible to regulate the temperature of the reaction mixture. A solution of 5 g. of benzoyl peroxide in 25 cc. of styrene was warmed to 90" over a period of twenty minutes. The temperature

+

+

+

+

(9) Shriner and Fuson, "Identification of Organic Compounds," John Wiley end Sons, Inc., New York, N. Y.,2nd edition, 1941, p. 112.

759

THEPOLYMERIZATION OF STYRENE

May, 1943

began to rise more rapidly so the mixture was removed from the steam-bath. The temperature continued to rise at an accelerated pace until it reached 140'; the reaction was accompanied by vigorous evolution of gas. The temperature immediately subsided and, after five more minutes at 86', the viscous mixture was cooled, dissolved in ether and poured into alcohol, yielding 14 g. of snow-white polymer. A solution of 1 g. in 10 cc. of ether was filtered into 50 cc. of QS% alcohol yielding 0.86 g. of polymer for analysis. A qualitative test by sodium fusion' gave negative results for nitrogen and for halogen. Anal. Calcd. for GH+202(Cd&)~: C, 90.61; H, 7.49; mol. wt., 1680. Pound: C, 0.63; H, 7.88; mol. wt., 1740 (visc.).' Polystyrene in Nitromethane.-A solution of 20 g. of styrene and 3 g. of benzoyl peroxide in 100 g. of nitromethane was heated on a steam-bath (95') for two hours. The mixture was cooled and poured into cold alcohol. A pale viscous oil separated. The mother liquor was decanted and the polymer was dissolved in 150 cc. of ether. When the solution was poured into 500 cc. of ice-cold alcohol, a white powdery precipitate separated. This was collected by filtration and washed with alcohol. The product weighed 10 to 12 g. A sample for analysis was prepared by dissolving in ether and filtering into three volumes of filtered icecold alcohol. Qualitative analyses for nitrogen by sodium fusion, using ferric thiocyanate or Prussian blue as a test,# were negative, as were qualitative tests for halogen. Anal. Calcd. for C&LCO~(CIH$I,: C, 90.51; H, 7.46; mol. wt., 1580. Found: C, 90.52; H, 7.46; mol. wt., 1400 (visc.).' Benzoyl Peroxide m d Nitromethane.-A solution of 20 g. (0.08 mole) of benzoyl peroxide in 30 cc. of nitromethane was heated at 100" for forty-eight hours. Extraction of the reaction mixture with sodium bicarbonate yielded 5 g. (0.04 mole) of benzoic acid. Fractional distillation of the organic layer yielded 1 g. (0.007 mole) of biphenyl, m. p. 70'.

Summnrv The elements of nitrobenzene and 2,4-dinitrochlorobenzene have been found in polystyrene prepared in the presence of small amounts of these aromatic nitro compounds. This indicates that the retarding effect of such compounds is due to their coupling with the active growing polymer chain. The effect must be due to the activation of the aromatic nucleus by the nitro groups since nitromethane has been found to be an inert solvent for the polymerization. URBANA, ILLINOIS

RECEIVEDFEBRUARY 1, 1943